Latest / The Index Podcast / Building Solana Apps with AI: Jordan O’Neill of Eitherway on No-Code Web3 Creation
Transcript
- 0:12SPEAKER_01: Hey everybody, welcome to the index.
- 0:14SPEAKER_01: I'm your host, Alex Cahaya, and today I'm joined by Jordan O'Neill, CMO and co-founder of Either Way.
- 0:21SPEAKER_01: He's got a background in digital marketing and web3 growth.
- 0:24SPEAKER_01: Jordan has helped crypto startup scale from through fundraising partnerships and go-to market strategy and now leads Eitherway's growth after bootstrapping the project with his own team.
- 0:34SPEAKER_01: And they have this really cool agenc product that they're building on Solana that we want to talk about.
- 0:39SPEAKER_01: But before we do that, thanks for being here.
- 0:42SPEAKER_01: And tell me more about what gets you out of bed in the morning.
- 0:46SPEAKER_00: Hey, Alex, good to meet.
- 0:48SPEAKER_00: And uh thanks for inviting us on the Enix show.
- 0:50SPEAKER_00: Uh big fan of what you guys are doing.
- 0:52SPEAKER_00: I've watched plenty of streams and see you guys endorse some of the best builders on Solana.
- 0:56SPEAKER_00: So we're excited to be here.
- 0:58SPEAKER_00: Um as you mentioned, I've been doing digital marketing for a number of years now.
- 1:02SPEAKER_00: I've been marketing crypto and incubation for startups, maybe better part of six years, I would say.
- 1:07SPEAKER_00: Um got involved with the either way team when we started building about 18 months ago.
- 1:12SPEAKER_00: Uh we're a registered C Corp in Delaware.
- 1:14SPEAKER_00: Uh, we bootstrapped 1.5 million to fund the venture.
- 1:18SPEAKER_00: Uh, we have a dev team of about 10, two designers, and then like a lot of strategic advisors to help us with like partnerships, BDs, KOL management, and onboarding.
- 1:27SPEAKER_00: So we have a very strong team.
- 1:29SPEAKER_00: Uh we launched our beta product, I would say, in December, and we had an event at a university in Morocco for healthcare students.
- 1:38SPEAKER_00: We had over 1200 signups and 800 apps built in a day, and we kind of realized then we're onto something here.
- 1:44SPEAKER_00: Uh, I would say what we started either way for was to develop the product that could simplify crypto complexities to make it easy for users to build any monetizable products in web three.
- 1:57SPEAKER_00: So I would say what we're doing categorically is similar to lovable, but what we do as a differentiator is we offer like web three premium infrastructures and uh tech stacks.
- 2:07SPEAKER_00: So we've a lot of great partners on that side to help us facilitate that, and you can prompt anything into a like ready-made production ready uh product.
- 2:17SPEAKER_01: And what's what's the name of the product?
- 2:18SPEAKER_01: Like what's the URL?
- 2:19SPEAKER_01: How do I how do I go check it out?
- 2:20SPEAKER_01: What's the URL?
- 2:22SPEAKER_00: E eitherway.
- 2:24SPEAKER_01: Very simple.ai, okay, cool.
- 2:27SPEAKER_01: Um, and so you guys have had this in beta since December, and you had this event in Morocco.
- 2:32SPEAKER_01: First of all, like what led you to Morocco?
- 2:34SPEAKER_01: That's kind of like a seems like a random place to me.
- 2:37SPEAKER_01: Um do you guys have some connection there with a team or like why did you do that there?
- 2:41SPEAKER_00: Yeah, so one of our uh one of our founders, his name is Mustafa.
- 2:45SPEAKER_00: He actually worked with KPMG and Larry Ev and stuff.
- 2:48SPEAKER_00: Uh he worked at the starting um setup in Chad and Kenya.
- 2:52SPEAKER_00: So he's very connected with governments in Africa.
- 2:55SPEAKER_00: So they actually invited us to do the event.
- 2:57SPEAKER_00: And like one of our advisors, his name is Nayer Abakar, he actually works with the UN and he invited us to, I'm not sure what the name of the university is, but it's a it's a big uh healthcare university.
- 3:09SPEAKER_00: We had a lot of different professional and professors there, students there.
- 3:13SPEAKER_00: It was more web 2 focused, like we had people making symptom AI apps and diagnose um your thing diagnose your uh symptoms online and stuff like that.
- 3:23SPEAKER_00: It was pretty cool, but we were trying to educate all of them then on like how to use USDC for payment rails and not just use credit card systems, not just use stripes.
- 3:31SPEAKER_00: So it was a good taste of like what the product is capable of, and then from there we had like a case study and social proof that kind of enticed some of the partners that we're working with on Web3 to actually take us seriously.
- 3:44SPEAKER_01: Yeah, I mean it makes sense.
- 3:45SPEAKER_01: Stable coins are way more important for emerging markets.
- 3:49SPEAKER_01: Um, you know, there's tons of adoption uh and I can see why that those rails would be really important.
- 3:56SPEAKER_01: I'm curious, can you show it to me?
- 3:58SPEAKER_01: Like, can we like do a quick screen share and we can vibe something out?
- 4:02SPEAKER_01: Like what what let's let's do something.
- 4:04SPEAKER_01: Let's do it.
- 4:05SPEAKER_00: So yeah, no, to be fair, um like our product, we kind of more so focus on like those type of demographic.
- 4:13SPEAKER_00: Can you see my screen?
- 4:14SPEAKER_00: Yeah.
- 4:15SPEAKER_00: Does it is that showing okay?
- 4:16SPEAKER_00: Do you referee okay?
- 4:18SPEAKER_01: So for people who are just listening and not watching on YouTube, he's got eitherway.ai slash chat up, and it says, don't just imagine it, launch it.
- 4:27SPEAKER_01: Um either way turns concepts into production ready apps across web, mobile, and web three.
- 4:31SPEAKER_01: No coding, no delays.
- 4:33SPEAKER_01: Just build, publish, and monetize either way.
- 4:35SPEAKER_01: Um so let's like what do we want to build?
- 4:37SPEAKER_01: What are we gonna build?
- 4:38SPEAKER_00: Anything.
- 4:40SPEAKER_01: Can you launch a perp stax?
- 4:41SPEAKER_01: Yeah, we could.
- 4:42SPEAKER_00: We'll use one of our aggregators.
- 4:43SPEAKER_00: So I'll go through some of the workflow so you can see how it works for everybody.
- 4:47SPEAKER_00: And I think what you covered on uh just there about like USDC accessibility and payment rails and stuff is like one of our main core focuses a product.
- 4:56SPEAKER_00: Like we're trying to get into demographics where like let's say they might be behind a little bit in the crypto race or behind in the AI race or whatever.
- 5:04SPEAKER_00: Uh, we want to make sure that like countries like Kenya and Chad and Nigeria and the Philippines and Morocco, they're not really forgotten about.
- 5:12SPEAKER_00: So we want to try and develop a system that we can add to like curriculums there or like have governmental deals with their countries to maybe offer a white label solution so that people that don't have the money to build apps and scale apps they can still build them for a cost-effective price and still have like good quality products.
- 5:29SPEAKER_00: So, a little bit about us, we don't do all of it ourselves, like everything is um configured in a certain way that our agent operates.
- 5:38SPEAKER_00: So we use like different infrastructure partners like Superbase for your back-end stuff.
- 5:43SPEAKER_00: Uh Stripe is for payment integrations, Helius, QuickNode are for RPC and validations.
- 5:48SPEAKER_00: Uh, this is our like cross-chain validation system and RPC.
- 5:53SPEAKER_00: Chain links for price, price articles and uh data fetching.
- 5:57SPEAKER_00: Moonpay is obviously on-ramp and off-ramp, but like through these integrations, like we can use open AI, you can connect your enthropy, connect your Netlify, your Verse L.
- 6:05SPEAKER_00: You connect all of these different platforms so that you can utilize them in your own app.
- 6:09SPEAKER_00: So if you want on-ramp or off-ramp, you can use Moonpay.
- 6:12SPEAKER_00: If you want to build like a launch pad, you can use Upstash.
- 6:15SPEAKER_00: If you want to use like anything for indexing, whether it be a prediction market, perps, chain link, um, and then helius then for validation, RPCs, price checking, price fetching, etc.
- 6:26SPEAKER_00: So, like as you can see here, I have tons of different apps that I've been building fully end-to-end.
- 6:31SPEAKER_00: And what we offer, like we can have you can add your own logos, you can add your own icons or images, whatever you want.
- 6:38SPEAKER_00: Cool.
- 6:38SPEAKER_00: You can upload files from Figma if you're working on something, you can import it here and use that front end, use that logic so it's not wasted.
- 6:45SPEAKER_00: It's very simple to connect your super base, just add your account.
- 6:49SPEAKER_00: We can use Stripe, we can use OpenAI, we can use Entropic.
- 6:52SPEAKER_00: Your back ends configured then.
- 6:53SPEAKER_00: And if you'd rather, so then we can use a different model toggle if you rather the output of Opus or Sonnet.
- 6:59SPEAKER_00: So I like Opus, does like pretty complex stuff, and then the key differentiator of what we do that other um let's say competitors don't do.
- 7:09SPEAKER_00: Like if you went to Lovable and you asked, build me a prediction market, it would send you to Calci and say, like, figure it out yourself.
- 7:17SPEAKER_00: Um, but with us, we have it all built in.
- 7:20SPEAKER_00: So you just toggle this, it goes from a web 2 agent to a web 3 agent, and then you can turn on like Metaplex for your DAS, you can turn on dialect for your blinks and swaps, and then if you say, like, let's build uh prediction market using D flow.
- 7:38SPEAKER_00: So D flow is one of our like uh recommended partners, and they use like Calci APIs um for their prediction markets.
- 7:45SPEAKER_00: So if I wanted to do like let's say stable coins and lending, I would use Camino.
- 7:49SPEAKER_00: If I want to use wallet authentication, we'd use Solflare.
- 7:53SPEAKER_00: If you want to do trading terminal or prediction markets, D flow.
- 7:57SPEAKER_00: Does that make sense?
- 7:59SPEAKER_01: Yeah, it does.
- 8:01SPEAKER_00: Um, so yeah, we've got a lot of infrastructure partners.
- 8:05SPEAKER_00: Uh we primarily focused on Solana to start, and then we partnered with Jupiter, Meteora, Radium, um, D Flow, Solflare, Camino, Bird Eye, like the list goes on and on.
- 8:18SPEAKER_00: And then we ran a hackathon with Super Team on Super Team Earn for Coliseum, and we had within 60 days like over 5,000 apps built.
- 8:28SPEAKER_00: We have 3,000 active users, and we've got about 2.5k successful deployments.
- 8:32SPEAKER_00: So that means like new map products released on Solana.
- 8:36SPEAKER_00: Yeah.
- 8:36SPEAKER_00: So we used that then as a case study and we leveraged some contacts.
- 8:40SPEAKER_00: We had a TRON and we're expanding to Tron Network, and we also have another chain that we're partnering with now to expand, and they've just started introducing us to uh infrastructure partners.
- 8:51SPEAKER_00: So our goal is to be like the interoperable solution for builders and just making it simple for anyone to build products in Web3.
- 9:00SPEAKER_01: Very cool.
- 9:01SPEAKER_01: And so right now it's like queued, the service is queued, it's kind of building it in the background, and then eventually it will deploy this somewhere.
- 9:08SPEAKER_01: Like, how does that work?
- 9:10SPEAKER_00: Yeah, so what happens is so as you can see, I turned on the web 3 agent.
- 9:14SPEAKER_00: So basically, the differential between the web 3 and the web 2 agent, so if you were to do like let's say uh scheduling or like a meeting planner or a fitness tracker or like something for the gym, you would you wouldn't want wallet authentication, you wouldn't need perps indexing, you wouldn't need price oracles, etc.
- 9:32SPEAKER_00: etc.
- 9:33SPEAKER_00: So you would just have the web 2 agent, and it is a little bit cheaper, let's say, but let's say for the web 3 agent, he is basically configured to execute things fully end-to-end.
- 9:44SPEAKER_00: So rather than just having placeholders, fancy UI, polished front-end work, it actually has all the features embedded, and instead of us building the tech stacks ourselves, our agent is configured to endpoints, where like let's say you you request the prediction market, our agent is configured to know D flow is where we go.
- 10:04SPEAKER_00: So you go to D flow, you go to their API, you pull the Calci API, you get your prediction markets, right?
- 10:09SPEAKER_00: So you get all your options.
- 10:10SPEAKER_00: Once you do your KYC, you're available to bet, etc.
- 10:13SPEAKER_00: etc.
- 10:14SPEAKER_00: Um, trading terminals, same thing, D flow.
- 10:16SPEAKER_00: Or if you want like on-ramp, off-ramp, Moonpay would be your option.
- 10:20SPEAKER_00: So all of these tech stacks that we've embedded and built into our infrastructure, they're what make building these apps end-to-end possible.
- 10:28SPEAKER_00: Because if we were to build the tech stack ourselves, it would be quite difficult, right?
- 10:32SPEAKER_01: Yeah, 100%.
- 10:33SPEAKER_01: And and it makes total sense.
- 10:34SPEAKER_01: Um what were some I'm just curious, like getting the age into you know, a lot of people that aren't in the weeds on how this works, like agentic workflows and building with agents, they don't realize how much fine-tuning is required, you know, and and how much knowledge goes into the systems and the harnesses and you know, all that.
- 10:54SPEAKER_01: I'm curious, like, what are some of the challenges you guys ran into if you can speak to it, you know, with that, like getting getting quality output?
- 11:02SPEAKER_00: Yeah, I would say like at the uh in the early stages, it was very difficult to get it to like ship anything other than like placeholders and templates.
- 11:11SPEAKER_00: Like we were we were all just shipping calculators for the longest time.
- 11:16SPEAKER_00: Because it was like it was an easy template to build, it was an app that you could configure easily, and the agent was able to execute it.
- 11:23SPEAKER_00: And then as time progressed, like we added like more polished UI, more polished front-end, more like different like templates and placeholders that could function on different like uh UIs.
- 11:33SPEAKER_00: Like if you were to build like an app, let's say for fitness, it would be similar to like a meeting scheduler.
- 11:38SPEAKER_00: So, like, and you can configure this to have like mobile optimization and stuff like that.
- 11:42SPEAKER_00: So I really think that the the difficult moment for us was like transitioning from just being a web 2, um, because it it goes from being a cloud wrapper then to being an actual fully agentic framework, and it has to have like we have to have integrations, right?
- 11:57SPEAKER_00: Because if we were to build a prediction market fully end-to-end from scratch, it wouldn't work.
- 12:01SPEAKER_00: You're gonna it's gonna cost way too much, it's months of dev time to figure out the configuration.
- 12:07SPEAKER_00: So we've we've had a very fortunate like networking opportunity going to different events like consensus and going to like Emerge and Miami Tech Week, where we've had like face-to-face meetings with a lot of these partners we work with.
- 12:19SPEAKER_00: Um, so I would say that the most difficult part was like getting it to like consistently um deliver like good products, and then once you get to that point where it's like it's showing a nice front end, it configures everything the right way you want it.
- 12:35SPEAKER_00: The UI, the design, it's about actual having functionality because like lovable, you have very nice placeholders, but minimal features, minimal functionality.
- 12:44SPEAKER_00: Having that end-to-end system where the app actually works, you can actually deploy an MVP, that was very difficult.
- 12:51SPEAKER_00: It was very, very difficult, and we're still improving, right?
- 12:53SPEAKER_00: Like we can be better on the UI, the front end.
- 12:56SPEAKER_00: I'm not sure why this is taking so long to load right now.
- 12:59SPEAKER_00: I might have to refresh it, or maybe we'll take a risk.
- 13:01SPEAKER_01: Let's try it.
- 13:02SPEAKER_00: Um let me try it again.
- 13:04SPEAKER_00: Let's refresh.
- 13:05SPEAKER_00: Um, usually this thing fires off.
- 13:08SPEAKER_00: Maybe there's an update going on, or I could just show you something else that I was working on.
- 13:12SPEAKER_00: Um I'll just get up so you can see how the chat works.
- 13:18SPEAKER_00: I'm not sure what this is, so it could be oh, I was building a GTA 6 game.
- 13:22SPEAKER_00: That's what I was doing.
- 13:23SPEAKER_00: I was trying I was trying to front run GTA 6 because I'm sick of waiting.
- 13:27SPEAKER_00: So this is um so yeah, I think like mainly the end-to-end execution was quite difficult, but now we're we're at a point where like our in our users can build like very sophisticated apps, very complex tech stacks.
- 13:43SPEAKER_00: You can use Camino, you can use um Radium, you can use Meteora, you can use D flow.
- 13:48SPEAKER_00: You don't have to just build a prediction market or trading terminal.
- 13:52SPEAKER_00: You can have an all-in-one encompassing app that you can build directly from either way.
- 13:56SPEAKER_00: So, like, for example, some of the the apps we had for the the hackathon, like there was some crazy stuff.
- 14:02SPEAKER_00: Like there was guys that done like polymarket betting for like football related, and they had like so many spreadsheets, so many lines of betting, they had so many different like um what are they called, like accumulators, they had like so many different spreads on it, and it was all football related, and it had like English leagues, Spanish leagues, Mexican leagues, like all different sort of leagues, and they'd used it, used QuickNode to build it, uh, which is crazy, and then obviously tapped into Calci API through uh D flow, done the KYC there.
- 14:32SPEAKER_00: You can also do the prediction markets through Polymarket, through Jupyter's API.
- 14:36SPEAKER_00: So um, yeah, it's been it's been a good journey.
- 14:39SPEAKER_01: Like let me ask you this like what when uh when a user comes to while while we wait for this thing to load, when when uh your users come, um what's like a best practice for prompting in your system so that they get a good result?
- 14:52SPEAKER_01: Is there anything that you need to do?
- 14:53SPEAKER_01: Like I'm I'm just I was looking at the screen here as uh you're scrolling through this workflow, and I see like a bunch of different prompts with like steps and things like that.
- 15:01SPEAKER_01: Are there some tips that you could give people that are gonna make them get better outputs in your system?
- 15:07SPEAKER_00: Yeah, so what I usually do is I like to plan it out firstly.
- 15:11SPEAKER_00: Um, and you can plan it out using either way, or you can plan it out using a different like AI model.
- 15:16SPEAKER_00: Uh, if you want to get into the Web3 side of things and you don't really know what each endpoint does, it can be good to ask your agent like what does Metaplex do?
- 15:24SPEAKER_00: What does MoonPay do?
- 15:26SPEAKER_00: What does Defoe offer?
- 15:27SPEAKER_00: And this agent will basically tell you exactly what endpoints that they can utilize and how they could best fit your app, let's say.
- 15:35SPEAKER_00: So they give you like the full rundown, our agent will give you the full rundown of like each utility and what their potential use case can be in your app, what the benefits are for your app, etc.
- 15:44SPEAKER_00: etc.
- 15:45SPEAKER_00: So yeah, I think um the best thing to do would be for me, I like using like Opus, I like using Grok, and then I get them to make me like a big prompt like this, where it's like complicated, it gives you the full rundown of exactly what you want, and then you just ship that into either way, and then it just builds, and then you go, you test your build, you test your preview screen, etc.
- 16:07SPEAKER_00: etc.
- 16:08SPEAKER_01: And um do you have any links to some of the stuff that like people built that like do you have an example of like uh you know that we could go look at, like an app?
- 16:16SPEAKER_00: Yeah, yeah, for sure.
- 16:17SPEAKER_00: Um let me shut down my screen.
- 16:18SPEAKER_00: So this is actually one that we had submitted to the Hackathon, which is really, really creative.
- 16:24SPEAKER_00: Um the Content Forge.
- 16:25SPEAKER_00: Content Forge.
- 16:26SPEAKER_00: So basically, what it is is it's an algorithm and an AI tool that basically fetches um everything on X, all different social medias.
- 16:35SPEAKER_00: It basically puts together like a growth strategy for companies to accelerate, like what content you should be posting, what are the trends, um, and basically it like it's an AI tool that caters and customizes customizes um like packages to what your product does.
- 16:53SPEAKER_00: So like if you have let's say like an app or like a trading terminal, let's say it would figure out like all the data and analytics across X and all different social medias of like how can you grow your terminal, how can you be competitive, etc.
- 17:07SPEAKER_00: etc.
- 17:08SPEAKER_00: As you can see, like competitive tracking, platform signal monitoring, on-chain intelligence, there's so many different features built into this, and he used either way to build all of this.
- 17:18SPEAKER_00: Like he's using D flow, he's using Moonpay, um, and I think they built their own AI algorithm for scraping, but like it's it's a fantastic tool for like automating content, um, for doing like specific ad campaigns for companies, um, and they they've done a terrific job.
- 17:36SPEAKER_00: Like it some of this stuff that was submitted through Super Team, I couldn't believe it.
- 17:39SPEAKER_01: No, this is super cool.
- 17:40SPEAKER_01: Yeah, this is super cool.
- 17:41SPEAKER_01: Yeah, I I use a skill all the time called Last 30 Days on my Clyde Code instance, and it like I use it for market research a lot.
- 17:48SPEAKER_01: Um, and it scrapes like X, TikTok, Instagram, YouTube, Reddit, and I was actually thinking about like turning it into a product with an interface like this.
- 17:57SPEAKER_01: Um more like web 2 focused without the like the web 3 stuff, kind of like the on-chain data and stuff like that.
- 18:04SPEAKER_01: And but this is a really good idea.
- 18:06SPEAKER_01: And I and I definitely have seen people who've seen me use it, their like eyes get super wide, like they're super excited about oh wow, this is very, very useful.
- 18:15SPEAKER_01: Um I'm curious, one thing that makes me ask or think of is you know, is there a way for me to download my source code so I can have it in my own repository and then add customizations there?
- 18:26SPEAKER_00: That's a great that's a great question.
- 18:27SPEAKER_00: So yeah, look, as I said, this thing is taking a bit too much time to load.
- 18:33SPEAKER_00: Maybe I'll go to a different one and it'll be different.
- 18:35SPEAKER_00: Maybe it's just for the game or whatever.
- 18:36SPEAKER_00: Back end is paused.
- 18:38SPEAKER_00: Yeah.
- 18:38SPEAKER_00: So I'll just show you here on one of these other ones, whatever I was building here.
- 18:43SPEAKER_00: So you can actually take your code and you can download your zip and you can have it just as raw code.
- 18:49SPEAKER_00: You own the code fully.
- 18:50SPEAKER_00: You can also deploy it on Netlify or sell GitHub, wherever you want.
- 18:54SPEAKER_00: So you own the code, you can export it from here.
- 18:57SPEAKER_01: Oh, so you can deploy it on GitHub, and that that will create a repo on GitHub.
- 19:00SPEAKER_01: Hey, finally.
- 19:02SPEAKER_01: There we go.
- 19:03SPEAKER_01: There we go.
- 19:04SPEAKER_00: So I was building a 3D city for uh app builders.
- 19:10SPEAKER_00: So if anyone was gaining traction on their app or like was getting crazy reviews, their building grows.
- 19:17SPEAKER_00: So basically the 3D city.
- 19:20SPEAKER_01: Are these apps built on your stack?
- 19:23SPEAKER_01: No, there were some apps that were these that we added.
- 19:26SPEAKER_00: So yeah, some of these were just mock-ups that we built ourselves just to have the configuration right, but like there's a lot of slap, as you can see, synth wave.
- 19:34SPEAKER_00: I've no idea what this is.
- 19:36SPEAKER_00: Cold pilot.
- 19:37SPEAKER_00: So there's a lot of placeholders, but some of these were basically just for monitoring what traction apps were getting.
- 19:44SPEAKER_01: No, it's super cool.
- 19:45SPEAKER_01: It's actually a really cool visualization.
- 19:46SPEAKER_01: You know, a lot of people talk about crypto as like an economy or like a country, right?
- 19:50SPEAKER_01: Like Solana is like its own country, its own economy.
- 19:53SPEAKER_01: And so you could put all the top Solana apps in here, and they would it would be like a city, you know?
- 19:59SPEAKER_01: And you could show like the transaction flows between apps and stuff as like little people or dots or something.
- 20:05SPEAKER_01: Like you have the dots moving around.
- 20:06SPEAKER_01: Um, that's actually really cool.
- 20:09SPEAKER_00: So yeah, it there would be no fluff as in you could see exactly what apps are getting tracted and who's getting what sort of reviews.
- 20:16SPEAKER_01: So how long so how long did this uh this take you to make?
- 20:19SPEAKER_00: This this particular build, this was like I would say like three to four days over a couple of sips.
- 20:26SPEAKER_01: Yeah, just hacking on it, just like prompting, adjusting, like everything it and stuff.
- 20:29SPEAKER_00: Everything done through here, as you can see.
- 20:31SPEAKER_00: Like as I said, I started off with like Grok and I done research onto like what could be a cool so I was looking at virtuals, right?
- 20:38SPEAKER_00: And I was like, what could be a flagship project for either way?
- 20:42SPEAKER_00: And I was like, oh my god, you know, like this could be a really cool idea and what they do with AI XBT and that agent, it was super engaging.
- 20:48SPEAKER_00: So this was like, oh well, you need to have a system.
- 20:51SPEAKER_00: So I asked Grok so many questions, and I said, like, look, do some research, figure out what an app building product needs to be the flagship.
- 20:57SPEAKER_00: So they came up with this idea, and it was like app city, 3D universe, you can monitor traction, like everything's cool.
- 21:04SPEAKER_00: But like, as I got into it, I was sending like screenshots and I was exporting the code, and I was like asking Claude and OpenAI to do like research.
- 21:13SPEAKER_00: I'm like, how can I improve it?
- 21:15SPEAKER_00: How can I improve it?
- 21:16SPEAKER_00: So when it came back to me then, I would have other prompts of like how I could get it to this point.
- 21:21SPEAKER_00: So I just like keep prompting, keep prompting.
- 21:24SPEAKER_00: There's absolutely no code needed at all.
- 21:28SPEAKER_00: This was another one I built, it was called Sally, but basically I used dialect and camino, and then I built an agent that basically tracks and fetches what's the best yield opportunity so that you can rotate, and then based on like your compounding interest, it gives you like better opportunities that pop up.
- 21:46SPEAKER_01: There you go.
- 21:47SPEAKER_01: Oh, bad gateway error.
- 21:48SPEAKER_00: Damn.
- 21:50SPEAKER_00: Maybe I have it shut down on my super base.
- 21:52SPEAKER_00: Oh, it's no, that's it.
- 21:54SPEAKER_01: That's all right.
- 21:55SPEAKER_00: So yeah, we had like all the different stocks plugged in, you have like different markets.
- 22:00SPEAKER_00: From Camino, cool, helium, uh, maple finance, and it basically just gives you all different options for yield, like stable coins, RWA, stocks, synthetics, precious metals, d pin, whatever you want.
- 22:12SPEAKER_00: And then there's like an agent built in that gives you recommendations of where you should start rotating.
- 22:18SPEAKER_01: Cool.
- 22:19SPEAKER_00: Yeah.
- 22:19SPEAKER_01: So that is super cool.
- 22:22SPEAKER_00: Yeah, and then like this this would be like we're we're developing our own open claw agent right now, um, which is like our own agent framework.
- 22:30SPEAKER_00: Uh, it's like open claw, I say it's like open claw because it's very similar as in like you have an autonomous agent assistant basically that builds after you, but it's specifically catered to either way, but it's fully open sourced.
- 22:43SPEAKER_00: Like you can have so instead of me developing complex prompts, I ask him for the plan, he lays out the plan, and then he gets to work.
- 22:50SPEAKER_00: And then if you're not good entin, it'll just like continuously improve your app.
- 22:56SPEAKER_01: That's super cool.
- 22:57SPEAKER_00: That's launching next week.
- 22:58SPEAKER_01: So I was can can I it can I use it on mobile?
- 23:01SPEAKER_01: Like, can I chat with it from Telegram and it will make things?
- 23:03SPEAKER_00: Yeah, yeah.
- 23:04SPEAKER_00: We have a bot.
- 23:05SPEAKER_00: So like directly on Telegram, you can see that on my screen?
- 23:09SPEAKER_00: Yep.
- 23:10SPEAKER_00: So we have an agent built on Telegram where you just basically connect your um your eitherway account.
- 23:16SPEAKER_00: Let's say we open it here, connect, go to either way, account linked.
- 23:22SPEAKER_00: As you can see, then you can configure whatever you want.
- 23:26SPEAKER_00: So your settings, your services, your Enclave variables, buying credits, whatever you need.
- 23:31SPEAKER_00: I'll go into a new app, I guess.
- 23:34SPEAKER_00: What do you want to name this app?
- 23:36SPEAKER_00: What what do we call it?
- 23:37SPEAKER_00: Index.
- 23:39SPEAKER_01: Let's do it.
- 23:41SPEAKER_00: So let's do Opus Web 2.
- 23:45SPEAKER_00: Because I was a little bit behind on time, I'm gonna build a a meeting planner.
- 23:51SPEAKER_00: A scheduling tool.
- 23:55SPEAKER_00: Okay, let's go.
- 24:01SPEAKER_00: Sorry, I need to give it a bit more context than that.
- 24:11SPEAKER_00: So yeah, then you can just like we can go away.
- 24:14SPEAKER_00: I know I can just leave this cook in my telegram chat, I can go off doing other activities or whatever, and when I come back, it'll be fully finished.
- 24:22SPEAKER_01: That's sick.
- 24:23SPEAKER_01: Well, we'll have to we'll have to post the link to that.
- 24:25SPEAKER_01: I'm I gotta think of something to build for the index show.
- 24:27SPEAKER_01: Yeah.
- 24:28SPEAKER_01: Something related, something related to what we do.
- 24:31SPEAKER_01: Um look at that.
- 24:33SPEAKER_01: It gives you so it's giving you the plan.
- 24:35SPEAKER_01: Yeah, and it's telling you you need to act like two credits used.
- 24:41SPEAKER_01: Ha, that's cool.
- 24:43SPEAKER_01: That's super cool, man.
- 24:45SPEAKER_01: Look, I really appreciate your time.
- 24:46SPEAKER_01: We're we're kind of getting towards the top of the show, and um I always ask this question, but what have I not asked you that I should have asked?
- 24:52SPEAKER_00: Um, I I guess like upcoming alpha, maybe.
- 24:56SPEAKER_00: Yeah, I would assume.
- 24:57SPEAKER_00: So we have this product right now.
- 25:01SPEAKER_00: We're fully revamping it.
- 25:02SPEAKER_00: Uh, we'll have like a new slick redesign coming out soon.
- 25:06SPEAKER_00: And this will be like a revamp that goes not only just for our product, but for all UIs and front ends that are going to be developed for either way.
- 25:14SPEAKER_00: So I would say we're gonna go from like a level five to six to like seven or eight, closer towards like lovable competition.
- 25:22SPEAKER_00: We've been more focused on back-end stuff, execution, web three integrations, and now we're going more so like scalability adoption.
- 25:31SPEAKER_00: Um we have this telegram agent that's fully developed.
- 25:35SPEAKER_00: We're now working on WhatsApp, Grok, and uh iMessage.
- 25:39SPEAKER_00: Uh, we want to make it an agent framework and network that can build like across so many different social media platforms so that you don't have to leave X.
- 25:47SPEAKER_00: You don't need to leave Telegram, you don't need to leave WhatsApp.
- 25:50SPEAKER_00: Um, what I'm really excited about is the open claw agent, uh, which is going to be like our own custom like either claw, let's say.
- 25:58SPEAKER_00: We're still discussing what the name of it should be, but it's gonna be a fully open sourced uh agent framework where you can basically have an autonomous assistant building all your apps fully end-to-end.
- 26:10SPEAKER_00: It maintains, it improves, it develops, it configures, and builds your whole app end-to-end autonomously, like fully automatic.
- 26:17SPEAKER_00: You don't need to touch it, you don't need to do all the coding, you don't need to do any prompting.
- 26:21SPEAKER_00: You give your uh agent the details, it checks in with where it's at, progress reports, and then you just keep building, keep building until you're happy with the output.
- 26:30SPEAKER_00: So I would say right now we're kind of in a phase where we have a lot of big updates coming, development updates in the next like week or two.
- 26:38SPEAKER_00: So I would say pay attention and keep an eye out.
- 26:41SPEAKER_00: Um I would like to say thank you for inviting me on this stream.
- 26:45SPEAKER_00: I know it was a little bit rushed, uh, a couple of things here and there that could have been a bit.
- 26:49SPEAKER_01: No, it's all good.
- 26:50SPEAKER_01: I and I love I love seeing the product, man.
- 26:52SPEAKER_01: Like, not many people actually even will do a demo live on the show.
- 26:56SPEAKER_01: And um, I know there was like a little bit of a hiccup there, but like that's totally normal for a startup.
- 27:01SPEAKER_01: I mean, this product is super slick.
- 27:03SPEAKER_01: Um, and it, you know, you guys are able to make real uh real applications um that have all these web three integrations in them, and I think there's something there, you know, and uh I'll definitely play around with it.
- 27:17SPEAKER_01: So if you're listening, go go check out eitherway.ai and start building some apps.